Case study · river flood risk study, Staffordshire
A village below a large water-supply reservoir flooded in February 2020. The hydrology was entirely my work. I built the 1D–2D hydraulic model jointly with an Environment Agency modelling colleague. I checked the model against residents’ own flood records, counted their mental-health harm alongside damage to homes and presented the results to the village. The protection delivered was a bund the farmers built.
Environment Agency and council references available on request.
Design standard: a flood with a 1% chance in any year (1% AEP), plus the 2080s climate allowance (+29% flow)
Real model output: the 2022 Flood Modeller 2D run (1% AEP flood + 29% climate allowance, 6 m grid; 149 ha at its peak), replayed from the model’s arrival times, not to time scale.
Animated map: in the 2022 model of a 1% AEP flood with a 29% climate allowance, water spreads over 79 hours to cover 149 hectares of the River Blithe and Trent floodplain at Hamstall Ridware.
Plan and appraise
What would stop the village flooding? The 2024 model tested two options against today’s situation, for a 1-in-100-year flood (1% AEP) with a 15% climate allowance. Drag the handle, or use the arrow keys, to compare them.
Baseline
Raised defence
Modelled depth, m
Change against the baseline (right of the handle)
In the baseline, water crosses the right bank into the village. Opening the second bridge arch gives no benefit upstream. The raised defence keeps the village dry and holds the water on farmland. Event: 1% AEP + 15% climate allowance.
Flood depths: Environment Agency 2024 model, which I built jointly with an Environment Agency modelling colleague; depth classes recovered from the report maps. Basemap © OpenStreetMap contributors (flood colours removed, farm labels blurred).
The whole study in five steps: the catchment, the hydrology, the model, the options and what was built. The detail behind each step follows.
Overview
Catchment → hydrology → 1D–2D model → options → outcome
Scroll through the five steps; in step 3, scrolling runs the flood hour by hour.Choose a step; in step 3, drag the hour slider to run the flood.
Four sub-catchments drain to the village
Design inflow above the reservoir, 1% AEP (hybrid)
2022 Flood Modeller 2D run: 1% AEP flood + 29% climate allowance, 6 m grid
1% AEP + 15% climate allowance (2024 model): baseline and raised defence
Raised defence
Baseline
Flood depths: Environment Agency 2024 model, built jointly by me and an Environment Agency modelling colleague. Basemap © OpenStreetMap contributors.
Benefit–cost ratio on one standard basis
My own self-initiated cost–benefit appraisal (not a formal Environment Agency submission): a simplified appraisal on a standard basis, using the damages from my 2022 study.
Step 1 of 5 · Catchment
Three-quarters of the catchment drains through Blithfield Reservoir, and at high flows water goes round the village gauge, so its record misses the biggest floods. I split the catchment at the reservoir and routed the upstream flood through it.
163 km²catchment to the village, about 75% of it controlled by the reservoir
Step 2 of 5 · Hydrology
I compared the gauge record with pooled statistics from similar catchments and the ReFH2 rainfall–runoff method, then built the design hydrographs from the most defensible parts of each: ReFH2 shapes scaled to pooled peaks, for each sub-catchment.
44.6 m³/spooled 1% AEP flow to the village, against 30.0 from the gauge record alone
Step 3 of 5 · 1D–2D model
The river and its bridge are modelled in one dimension, the floodplain and village in two. The hydrology was entirely my work. I built the 1D–2D hydraulic model jointly with an Environment Agency modelling colleague. I checked it against residents’ records of the February 2020 flood. Keep scrolling to run the 2022 model through the event.
149 haflooded by hour 79 in the 2022 run (1% AEP flood with a 29% climate allowance)
Step 4 of 5 · Options
The 2024 model (Flood Modeller–TUFLOW, 5 m) tested opening the second bridge arch and raising the right-bank defence. Opening the arch gave no benefit upstream. The raised defence keeps the village dry and holds the water on farmland.
Village drywith the raised right-bank defence, for every event modelled up to 0.1% AEP
Step 5 of 5 · Outcome
In my own cost–benefit appraisal (self-initiated, not a formal Environment Agency submission), on one standard basis, an earth bund on the defence line has by far the highest benefit–cost ratio of the options appraised. What was delivered was an earth bund built by the farmers, through partnership with landowners and the village.
4.6benefit–cost ratio of an earth bund (3.3–7.0 across unit rates), against 1.6 for a wall and 1.0 for an embankment
My flood estimation record v4 (2025); the Environment Agency 2024 model, built jointly by me and an Environment Agency modelling colleague; the 2022 Flood Modeller 2D run, rendered from its arrival-time and maximum-depth grids; my own self-initiated standard-basis appraisal (2026). Basemap © OpenStreetMap contributors.
Before the flood: understanding risk
Where the water comes from, and why the usual methods did not fit. Most of the catchment is controlled by a large reservoir, and the village gauge cannot measure the largest floods.
Each sub-catchment has its own design flood; the reservoir catchment’s flood is routed through Blithfield Reservoir before it joins the others.
Traced from my flood estimation record (2025). Schematic, not to survey accuracy.
How often do big floods happen here? The village gauge gives fifty years of each year’s highest flow (1959/60 to 2020/21), fitted with a Gumbel distribution, a standard curve for yearly maxima. Above the bypass flow the record flattens, because water goes round the weir rather than through it.
Dots: annual maxima at Gringorten plotting positions; hover for the water year. Lines: Gumbel fit to the gauge record, and the pooled estimate for the full catchment.
Gauge data: NRFA station 28002 © UKCEH; Environment Agency. My analysis (2022, updated 2025).
Three standard ways of estimating a rare flood gave different answers, so I combined the most defensible parts of each: the gauge record, pooled statistics from similar catchments (WINFAP) and the ReFH2 rainfall–runoff method.
Full-catchment (lumped) estimates, used as a check. The design flows are the final hybrid values for each sub-catchment, listed in the table.
My flood estimation record v4 (March 2025). Methods: Flood Estimation Handbook statistical (single site and pooled, WINFAP) and ReFH2.
A 2D model needs the whole flood, not just its peak. These are the inflows that drive the model, for each return period and with each climate allowance.
Climate-allowance lines are derived by scaling the base hydrograph by (1 + allowance), the way peak-flow allowances are applied; they are dashed and labelled “derived”. The upstream inflow is shown before routing through the reservoir.
ReFH2 and pooled statistical analysis, flood estimation record v4 (2025); sub-catchment series digitised from the December 2024 technical report (approximate).
Water that spreads onto the fields upstream of the village does not arrive at once. In the 2D model the floodplain lowers and delays the flood peak.
Approximate, digitised from the 2022 model output: a peak of about 34 m³/s in becomes about 19 m³/s out, some 31 hours later. The shaded gap is water held on the floodplain.
Flood Modeller 1D–2D model (2022), built jointly by me and an Environment Agency modelling colleague; values read from the model’s flow plot.
A computer model shows where the water goes, field by field and street by street. The river channel and its bridge are modelled in one dimension; the floodplain and village in two. The hydrology was entirely my work. I built the 1D–2D hydraulic model jointly with an Environment Agency modelling colleague.
Upstream inflow: routed reservoir outflow plus tributaries. Downstream boundary: a fixed level on the River Trent.
Model extent traced from the 2024 modelling report; roads and water © OpenStreetMap contributors. Schematic.
I build full 2D hydrodynamic models at 10–30 m resolution. I used TUFLOW and Flood Modeller 2D here; HEC-RAS 2D and SFINCS are other options for river and coastal work. At Hamstall the grid is finer, 5 m, because the decision turns on a single bank line beside a village.
Homes reported flooded by residents in a community questionnaire, against homes flooded in the model, on three public roads (labelled generically). Hover for response notes.
Community flood questionnaire and the 2022 Flood Modeller model; February 2020 flood (gauge peak 20.8 m³/s).
Cells wet by each hour, coloured by the deepest water they eventually reach. The faint area is the full extent at hour 79. Deep water to the south is the River Trent floodplain inside the model domain.
Event: 1% AEP + 29% climate allowance, March 2022 Flood Modeller 2D run (6 m grid). Rendered exactly from my raw time-of-first-flooding and maximum-depth grids; British National Grid 410000–412418 E, 317004–320088 N.
Learn and adapt: the future will differ from the past
What changes as the climate warms? The 2024 model ran the 1-in-100-year flood (1% AEP) with three climate allowances, and a 0.1% AEP flood for context. Step through them to see how many buildings flood, and how deep.
Buildings listed as flooded in the model, and the depth of water at them. The gauge shows the maximum depth at any building.
Environment Agency 2024 model (Flood Modeller–TUFLOW, 5 m), built jointly by me and an Environment Agency modelling colleague.
Plan and appraise
I also carried out my own cost–benefit appraisal of the options (a self-initiated appraisal, not a formal Environment Agency submission). It weighs what each defence would cost over 100 years against the flood damage it would avoid. All three options use one standard basis and the damages from my 2022 study.
Select an option for how it was appraised. All three options are on one standard basis, so their benefit–cost ratios can be compared directly. GBP, 2021–22 prices.
My own self-initiated standard-basis appraisal (2026), not a formal Environment Agency submission, using the event damages from my 2022 study, Environment Agency fluvial defence cost evidence (SC080039), FCERM discount rates and 60% optimism bias.
Deliver
The appraisal and the 2024 model both pointed to holding the line on the right bank. What was delivered was an earth bund built by the farmers, through partnership with landowners and the village.
On one consistent basis, the earth bund has by far the highest benefit–cost ratio of the options appraised: about 4.6, against 1.6 for the wall and 1.0 for the embankment. The bund row in Fig. 11 appraises an earth bund of that kind on the modelled defence line; the farmers’ bund itself has no recorded dimensions or cost, so that row describes the option, not the as-built bank.
The modelling evidence did the work it should: it showed where water entered the village, what each option would change, and that a defence on the right-bank line was worth having. The solution was then delivered through partnership: the farmers built the bund themselves, working with landowners and the village, supported by the Environment Agency’s modelling evidence.
ADB’s Resilient River Basin Initiative looks at the modelling, monitoring, forecasting and decision-support tools that river basins depend on. Assessing any of them comes down to four questions. Here are Hamstall’s answers.
| Question | At Hamstall Ridware |
|---|---|
| What tools were used? | The village gauge record; Flood Estimation Handbook statistical methods with WINFAP pooling and the ReFH2 rainfall–runoff model; routing through the reservoir; a linked Flood Modeller 1D and TUFLOW 2D model on 1 m LiDAR with a surveyed channel; Multi-Coloured Manual damages and a cost–benefit appraisal. |
| Were they fit for the decision? | Not on their own. The gauge is bypassed in floods, so single-site statistics came out a third below pooled estimates. Splitting the catchment at the reservoir, reconciling three methods and checking the model against the community’s flood records made the evidence fit to choose between options. |
| What data and capacity did they need? | LiDAR, a channel and structures survey, the reservoir operator’s data, residents’ records of a real flood, and a modeller able to build, run and document the chain for quality assurance. |
| How could they be strengthened? | Measure high flows at the gauge, or rate a bypass-proof section; keep model runs and routing workbooks together as a maintained model; carry the 2024 results into an updated appraisal; and link modelled levels to a local flood-warning trigger. |
Environment Agency and council references available on request.
Credits: Environment Agency 2024 model (built jointly by me and an Environment Agency modelling colleague); basemap © OpenStreetMap contributors (flood colours removed, farm labels blurred); NRFA gauge 28002 data © UKCEH; Environment Agency LiDAR (Open Government Licence).
Property names and addresses are withheld; verification counts are grouped by public road.